I think they should be treated the same, to increase our fun and strategy, while reducing our MM. However, I do not think, I am not saying, nor am I evangalizing that Star Bases should be Mercury sized creations. They should not. I can see that Star Bases may grow to be the size of England or Germany, with sufficent time and technology and resources, but not something as big as, say, Mars or even the tiniest of our "worlds", Pluto. In my mind, structures that big would have a much different set of engineering difficulties then a set of mutually cooperating, politically/economically aligned, multiple facilities which maintain their own relative floating placements via station engines. Remember, a Star Base doesn't have to be one "building". It could just as easily be a chain of mixed facilities (or even mixed complexes), sharing common local regulations and participating in a local administration to oversee that "business" continues on. The main limiter to how many such facilities you could keep near each other would be the economics of maintaining the structure before the collected mass would want to collect together around their common center of gravity (ie, the cost of keeping them from falling together). That's the real point of differentiation for me. A star base can be just a small refueling depot with a big warehouse of spare parts located a mere 2km away from the main administration and fueling facility. Or it could be a huge, sprawling affair that is a collection of mega-hotels, theme parks, casinoes, night clubs, zoos, theatres, mega-malls, gift shops, and resteraunts, with all the local warehouse, shipping and receiveing, and all the ship and space structure maintenence facilities maintain it and to keep it all running. An artificial world will be designed so that it plans on, depends, and takes advantage of, being a world with its own mass. Whether thats something that is merely 200km radius or something 12000km radius.
As for actual material, in reality, according to current astronomy, we "know" that there are literally worlds of all sizes and types just flying about through space, tossed out from their birth system. The current best of our knowledge puts that at about 6 times the number of stars in existance. That's a lot of material. For a star faring empire, there is plenty of raw material which can be harvested or moved to a more convienent spot, for world construction. For simple star base construction, you wouldn't need to go to such extremes. The formation of a solar system is a messy thing. There's lots of asteriods, comets, and planetoids left about, even after more then half of what was born in that system is ejected or destroyed (current best planetary science). Asteriods of usable size and material can be collected and hollowed out or simply reshaped to whatever is convenient. Comets can be captured and harvested for useful material. Remember also that material can be harvested not only from deep in a gravity well (like, Earth), but from shallower gravity well sources (meaning, cheaper). Indeed, our moon itself would make a cheap place to harvest everything you need to make cement according to various NASA studies. It would be much cheaper then shipping it up from Earth, for instance. That material could then be used in space construction, both in orbit or luna-side (proposals made by NASA, JAXA, and India). Harvesting iron, carbon, hydrogen, and oxygen from local resources in our solar system will not be a challenge for us, merely time consuming. For a civ with more advanced stellar travel then our own, it should take a lot less time for them to get around, which makes doing so much more feasible and economic.
Again, I'm *not* saying that a Star Base would be the size of a planet. It should never be such (for, most importantly, game play reasons). But consider, we don't use any significant portion of planets under the current GC2 tile plan. Each tile is just a complex with its own support stuctures (like company mining town). So it should be easy to place that same complex in space. Just more expensive to build and maintain. Consider, if we wanted to build a new Los Alamos like research mega-plex on a remote island in the South Pacific, we could do it. But, it would cost much more then building it out in Kansas.